A tetrahedron-based endmember selection approach for urban impervious surface mapping.

Wei Wang,Minhe Ji,Junpeng Zhai,Xinfeng Yao,Ben Bond-Lamberty

doi:10.1371/journal.pone.0093479

Abstract

The pixel purity index (PPI) and two-dimensional (2-D) scatter plots are two popular techniques for endmember extraction in remote sensing spectral mixture analysis, yet both suffer from one major drawback, that is, the selection of a final set of endmembers has to endure a cumbersome process of iterative visual inspection and human intervention, especially when a spectrally-complex urban scene is involved. Within the conceptual framework of a V-H-L-S (vegetation-high albedo-low albedo-soil) model, which is expanded from the classic V-I-S (vegetation-impervious surface-soil) model, a tetrahedron-based endmember selection approach combined with a multi-objective optimization genetic algorithm (MOGA) was designed to identify urban endmembers from multispectral imagery. The tetrahedron defining the enclosing volume of MNF-transformed pixels in a three-dimensional (3-D) space was algorithmically sought, so that the tetrahedral vertices can ideally match the four components of the adopted model. A case study with Landsat Enhanced Thematic Mapper Plus (ETM+) satellite imagery in Shanghai, China was conducted to verify the validity of the method. The method performance was compared with those of the traditional PPI and 2-D scatter plots approaches. The results indicated that the tetrahedron-based endmember selection approach performed better in both accuracy and ease of identification for urban surface endmembers owing to the 3-D visualization analysis and use of the MOGA.

Highlights

Urban composition exhibits a high degree of spatial and temporal heterogeneity, which is typically characterized by sharp boundaries and frequent variation of surface materials over space and by their changes mostly as a result of human activities over time [1]
The comparison was performed quantitatively with the measure of root mean square error (RMSE) (see eq (6)) between the observed and predicted values, hereby aiming to evaluate the overall accuracy of an impervious surface fraction map resulting from the linear spectral mixture analysis (LSMA) model
For all three endmember extraction methods involved in the analysis, the RMSE of the entire map was 0.014 for the 3-D method, 0.016 for 2-D scatter plots, and 0.021 for pixel purity index (PPI)

Summary

Introduction

Urban composition exhibits a high degree of spatial and temporal heterogeneity, which is typically characterized by sharp boundaries and frequent variation of surface materials over space and by their changes mostly as a result of human activities over time [1]. Mapping the dynamics of urban composition, the impervious surface which has emerged as an indicator of urbanization, is essential for understanding human-environment interactions, and for better urban planning and management decisions [2,3]. In recent years, this task has been increasingly performed using remote sensing technology, as it is deemed an effective tool to estimate impervious surface for large areas with relatively low costs and high suitability [4]. This issue has triggered many studies that are aimed to explore ways of extracting urban surface components at the sub-pixel level [7,8,12]

Objectives

Methods

Results

Discussion

Conclusion